#!/usr/bin/perl
# Copyright (c) 2006 Eclipse Foundation, made available under EPL v1.0
# Contributors Ward Cunningham, Bjorn Freeman-Benson


use strict;

print "Content-type: text/html\n\n";

my ($project)=$ENV{QUERY_STRING} =~ /project=([a-z0-9.]+)/;
$project = 'technology.dash' unless $project;
my ($count) = $ENV{QUERY_STRING} =~ /count=(\d+)/;
$count = 200 unless $count;
my ($days) = $ENV{QUERY_STRING} =~ /days=(\d+)/;
$count = 100 unless $count;

my %words;
my %breakdown;

my $lastmonth;

print "<table cellpadding=10><tr><td valign=top>";
projects();
print "<td valign=top>";
tags();
print "</table>\n";

sub trouble {
	print "Trouble: $_[0]\n";
	die $_[0];
}
sub head {
	my ($title) = @_;
        print <<;
                <table width=100%><tr><td bgcolor=#88ff88 align=center>
                <font color=white size=+2>
                $title
                </font>
                </table>

}
sub tags {
	my %months;

	open (F, "uniq/$project") or trouble ("$project: $!");
	while ($_ = <F>) {
		s/(\d\d\d\d\d\d)\d\d//;
		my $mo = $1;
		$months{$mo}++;
		my @words = /(\w+)/g;
		for (@words) {
			$words{$_}++;
			$breakdown{$mo}{$_}++;
		}
	}

	my @words =  keys %words;
	my $words = @words;
	my $frequent = $words < $count ? $words : $count;
	my @frequent = (sort {$words{$b} <=> $words{$a}} @words)[0..($frequent-1)];
	my ($max, $min) = ($words{$frequent[0]}, $words{$frequent[$#frequent]});

	head ("Lifetime");
	for (sort @frequent) {
		my $n = $words{$_};
		my $ratio = ($n+1 - $min) / ($max - $min);
		my $size = int(log($ratio)+7);
		print "<font size=$size>$_</font>\n";
	}

	my @months = keys %months;
	my $months = @months;

	for my $m (sort @months) {
		my %mwords = %{$breakdown{$m}};
		my @mwords = keys %mwords;
		my $mwords = @mwords;
		for (@mwords) {
			$mwords{$_} = $mwords{$_}*$months - $words{$_};
		}
		my $frequent = $mwords < $count ? $mwords : $count;
		my @frequent = (sort {$mwords{$b} <=> $mwords{$a}} @mwords)[0..$frequent-1];
		my ($max, $min) = ($mwords{$frequent[0]}, $mwords{$frequent[$#frequent]});

                my @mon = qw(Nil Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec);
                $m =~ /(....)(..)/;
                head("$mon[$2] $1");

		for (sort @frequent) {
			my $n = $mwords{$_};
			my $ratio = ($n - $min + 1) / ($max - $min + 1);
			my $size = int(log($ratio)+7);
			print "<font size=$size>$_</font>\n";
		}
	}

	print <<;
		<br><br><table width=400><tr><td><font color=gray size=-1>
		These $frequent words are the most frequent of
		$words words found on $. commit 
		messages recorded over $months months by the 
		$project project.
		See <a href="uniq/$project">raw data</a> used by this script.
		</font></table>

}

sub projects {
	opendir (D, 'uniq');
	my @projects = (grep !/^\./, readdir(D));
	for (@projects) {
		my $size = "size=+2" if $_ eq $project;
		print "<font $size><a href=http:tags.cgi?project=$_&count=$count>$_</a></font><br>\n";
	}
}